Electrical & mechanical and transmission & distribution specialty contractor · data-center / grid build-out exposure · Bismarck, ND
Damodaran FCFF (ginzu standard) · 5-yr margin-normalized · pro forma Epsilon · 10-yr + stable · reverse DCF · valued 2026-10-15 (refresh of 9/29 build)
OVERVALUEDRISK: MEDIUM-HIGH
Price $108.94 · Base $88.10 · Price is +24% above base
What they do
Everus is the old construction-services arm of MDU Resources, spun off as a standalone company on 10/31/2024. It does not build whole buildings. It is the specialty subcontractor that wires, pipes and powers them, and it builds and maintains the power lines that feed them. It has about 9,400 employees, mostly union electricians and linemen, working out of offices across the western and central US.
Two segments.Electrical & Mechanical (E&M), roughly three-quarters of revenue: electrical and low-voltage wiring, HVAC and mechanical piping, and fire suppression for data centers, hospitals, hotels and casinos, manufacturing, schools and government buildings. Data centers are now the largest slice of backlog. Transmission & Distribution (T&D), roughly a quarter: builds and maintains overhead and underground electric, gas and communications lines for utilities, plus substation work, and rents line-stringing equipment. The 9/1/26 Epsilon acquisition (~$250M revenue) adds off-site modular electrical and mechanical prefabrication. Segment mix is approximate, from FY25 disclosures.
How it makes money. It bids on projects and earns a construction margin on labor and materials. About 52% of revenue is fixed-price, where cost overruns come out of Everus's margin. The rest is cost-plus or time-and-materials, which is lower risk. Revenue follows backlog ($4.55B at June, +53% Y/Y). The business needs little capital: capex runs about 1–2% of revenue. The swing factor in cash flow is working capital, in particular customer prepayments.
Key sensitivities: hyperscaler and data-center capex (the top customer is ~17% of revenue), how accurately it bids fixed-price jobs, the supply and productivity of skilled electricians, utility grid capex budgets, and interest rates through commercial construction starts.
Price (10/15 intraday, −4%)
$108.94
Base intrinsic value
$88.10
Bear → Bull · prob-wtd (25/50/25)
$54 – $135 · $91
Key driver
Data-center growth runway ±5 pts/yr = $69–$112
The Take — Still not cheap. The stock fell 7%, the value didn't move, and it's still pricing a bull case.
Nothing fundamental has changed in 16 days. There's been no new filing, no guidance change and no backlog update since the 9/29 build. The only events were Epsilon closing on 9/1 (already in the model), Cantor cutting its target to $152 (Neutral) and DA Davidson starting at Buy, $168. The model is rerun with the same inputs, and base value is unchanged at $88.10. The price went from $116.68 to $108.94, so the premium to base fell from +32% to +24%. That's still above the +15% OVERVALUED line.
The business is excellent, but the price assumes it stays at its peak. TTM revenue is $4.27B (+26%), EBIT is $331M (7.75% margin), backlog is $4.55B (+53%), and after-tax ROIC is ~27%. FY26 guidance was raised twice, to $4.5–4.7B revenue and $410–425M EBITDA. But the 2021–25 average EBIT margin is 6.8%. At that margin, TTM earning power is $291M, not $331M. Management expects H2 margins to normalize, and its long-run organic growth target is 5–7%.
The reverse DCF shows how far the bar has dropped. On the base growth path (18% → 7% over five years), $109 needs a 9.1% EBIT margin forever, down from 9.75% at $117. That still matches Q2'26, the best quarter in the company's history. On the bull path (24% → 9%), the price needs only 7.7%, which is today's TTM margin. Put plainly, $109 is a bet that data-center demand stays near peak for five more years. That's possible. It isn't a margin of safety.
What changed vs the 9/29 build: price −6.6%. Base value is unchanged at $88.10 with the same inputs: riskfree 5.20%, WACC 9.5%→9.0%, 7.5% target margin, Epsilon pro forma. Bull is $135 (unchanged). The bear rebuilt at $54 versus $55 previously (a small difference from how the stable WACC was reconstructed). The verdict is the same, and the gap has narrowed, not closed.
Call: AVOID at $109. Accumulate below ~$80. Starter at ~$88 (base) only if Q3 confirms backlog above $4.5B and margins at or above 7.5%. Holders should trim into $115–125 or write covered calls; don't add. Next catalyst: Q3'26 earnings, expected early-to-mid November (StockAnalysis estimates 11/3; TipRanks lists 11/11; not yet announced). Options: for a bearish lean, use a defined-risk bear call spread above the 52-week overhang (e.g. short $130 / long $140 calls, Dec/Jan expiry, sized to the debit-at-risk). Patient buyers can sell cash-secured puts at the $80 strike into the post-earnings IV, which gets you paid to wait for the entry level. No naked calls.
Risk Profile — Medium-High
HIGH
Data-center capex cycle / concentration. Data centers are the largest slice of backlog. The top customer is ~17% of FY25 revenue and the top 10 are 43%. A hyperscaler pause hits growth and overhead absorption at the same time. Long the driver.
HIGH
Fixed-price execution. 52% of revenue is fixed-price, and margins are at a record. One bad job wipes out a quarter. ±100bp on the target margin moves value ±15%.
MED
Rates. The 10Y is above 5% (5.2% riskfree). WACC ±100bp moves value to $73–$109. Higher rates also slow non-hyperscale commercial and institutional starts. Short rates.
MED
Working-capital reversal. H1 FCF of $167M was flattered by a $114M build in customer prepayments (contract liabilities $419M). Management has flagged that this reverses. Normalized FCF is ~$100–130M/yr.
MED
Labor. Electrician scarcity is a tailwind on pricing but a ceiling on growth. Wage inflation on fixed-price jobs squeezes margins. Long union-labor productivity.
LOW
M&A / leverage creep. $148M for SE&M plus $295M for Epsilon (~10.7x EBITDA). Pro forma net debt is ~$415M (~1.0x EBITDA). Manageable.
For a contractor, the "inventory" is backlog plus skilled labor. The real bet is how long the data-center and grid build-out lasts, at record fixed-price margins. You are long hyperscaler and utility capex, long electrician productivity, and short rates. There's no commodity or FX line, but one cyclical driver sets both growth and margin.
Scenario values per share vs price
Bear: growth 8%→4%, margin 6.5%, WACC 10%→9.5%. Base: 18/14/11/9/7% then fading to 3%, margin 7.82%→7.5%, WACC 9.5%→9.0%. Bull: 24/20/16/12/9%, margin 8.5%, g 3.5%, stable ROIC 20%. No blue-sky case: there's no plausible structural margin break-out for an E&C sub. Weights 25/50/25 = $91.19.
Driver tornado: value/share vs base $88.10
The widest bar is still the near-term growth runway ($69–$112), then WACC and margin. This is a bet on how long the data-center / grid build-out lasts and whether fixed-price margins hold at record levels — not a multiple call. Terminal assumptions barely matter: TV is 62% of operating value, inside the 70% guardrail.
Master-driver grid: value/share by growth path × steady-state EBIT margin
Y1–5 growth ↓ / margin →
6.5%
7.0%
7.5%
8.0%
8.5%
9.0%
9.5%
-6 pts/yr
$57
$62
$66
$71
$76
$80
$85
-3 pts/yr
$65
$71
$76
$82
$88
$93
$99
+0 pts/yr (base 18→7%)
$75
$81
$88
$95
$102
$108
$115
+3 pts/yr
$86
$94
$102
$110
$118
$126
$134
+6 pts/yr
$99
$108
$118
$127
$137
$146
$156
Green = ≥5% above price $108.94; yellow = within ±5%; orange = up to 30% below; red = more than 30% below. Outlined = base. On the base growth path the price needs a ~9.0% margin held forever; at +6 pts/yr of extra growth (the bull path) it needs ~7.7% — i.e. today's TTM margin. That is what $109 is pricing.
Sensitivity: value/share by WACC × target EBIT margin (base growth)
WACC ↓ / margin →
6.5%
7.0%
7.5%
8.0%
8.5%
9.0%
9.5%
8.5% → 8.0%
$93
$101
$109
$117
$125
$133
$142
9.0% → 8.5%
$83
$90
$98
$105
$112
$120
$127
9.5% → 9.0%
$75
$81
$88
$95
$102
$108
$115
10.0% → 9.5%
$68
$74
$80
$86
$92
$99
$105
10.5% → 10.0%
$62
$68
$73
$79
$85
$90
$96
Green cells require an 8.5%+ permanent margin at a sub-9.5% WACC, or 9.5% at base WACC. Best-ever quarter: 9.1% (Q2'26).
Sensitivity: value/share by WACC × terminal growth
WACC (hi → stable) ↓ / g →
2.0%
2.5%
3.0%
8.5% → 8.0%
$99
$104
$109
9.0% → 8.5%
$90
$93
$98
9.5% → 9.0%
$82
$85
$88
10.0% → 9.5%
$75
$77
$80
10.5% → 10.0%
$69
$71
$73
Same color rules. Spec cap for cyclicals is 2.5%; the base keeps the prior build's 3.0% for consistency (2.5% = $85). No cell reaches the price.
Pale bars: 2021–22 are estimates (MDU Construction Services segment NI grossed up for tax and interest). 2026G is guide-implied (midpoint revenue ~$4.6B, EBIT ~$375M), not reported.
Normalization worksheet ($M)
Year
Revenue
GAAP EBIT
Acq. amort / one-offs
Clean EBIT
Margin
R&D
2021 (est.)
2,050
150.0
—
150.0
7.3%
none
2022 (est.)
2,700
171.0
—
171.0
6.3%
none
2023
2,854
190.5
n/d
190.5
6.7%
none
2024
2,850
189.9
n/d
189.9
6.7%
none
2025
3,746
264.8
n/d
264.8
7.1%
none
TTM Jun-26
4,267
331.0
n/d
331.0
7.8%
none
5-yr avg margin (2021–25)
6.8%
× TTM revenue $4,267M = normalized EBIT
$291M
vs TTM $331M
GAAP→clean bridge: acquired-intangible amortization and one-offs were not separately disclosed in the sources available for this refresh (the SEC filing index returned no ECG filings). They are small for this business (few acquired intangibles before the 2026 SE&M/Epsilon deals). Clean EBIT = GAAP EBIT, consistent with the 9/29 build. SBC stays expensed. No R&D, so no capitalization. This is margin normalization, not dollar-EBIT averaging, which would badly understate a firm growing ~30%. The base starts at TTM 7.82% and glides to 7.5%: above the 6.8% 5-yr average, crediting the structural mix shift to data centers, and below the 9.1% peak quarter.